EN

Identifying the 6 leads of a three-phase motor

Letters and polarity with reduced-voltage AC

The problem

You have a motor with six loose, unmarked leads. A multimeter tells you which two leads form each coil, but not which is the start and which is the end. If one coil ends up reversed, the motor starts badly or overheats: that is the fault you can see in Rotating field with the invert-phase buttons. The method is to induce voltage from one coil into the other two using reduced AC. This site names the coils U-X, V-Y and W-Z; many current terminal boards use U1-U2, V1-V2 and W1-W2.

What you need

Voltage. 10 to 12 times lower than the coil voltage. On a 230/400 V motor the coil is 230 V, so about 19 to 23 V. On a 400/690 V motor, about 33 to 40 V.

Power. The larger the motor, the more power you need. With the cage rotor stopped, the coil behaves like a transformer with a short-circuited secondary: at one eleventh of the voltage it can draw around half its rated current. A small source collapses and the reading comes out weak. As a guide, here are values we calculated, not measured, for 4-pole direct-on-line motors:

Motor (kW)Source, approximate (VA)
1.1≈ 35
4≈ 100
11≈ 250
22≈ 500
45≈ 1000
90≈ 2000

Check the current with the clamp as soon as you apply voltage: it must not exceed the coil's rated current.

Safety

Step 1: separate the three coils

With the multimeter on continuity, find which two leads belong together and mark them as a pair. You should get three pairs with no continuity between them and similar resistances.

If you do not get three pairs, stop: it may be another type of motor, such as a Dahlander or a two-speed motor. This method is for three independent coils.

Step 2: first test

  1. Feed the first pair. The direction does not matter: we will call it U-X, even though we do not yet know which lead is which.
  2. Join one lead of the second pair to one lead of the third. Put the voltmeter across the two free leads.
  3. Apply voltage and read.
  4. If there is a clear voltage, the joint is end to start. Mark the free lead of the second pair V, its joined lead Y, the joined lead of the third pair W and its free lead Z.
  5. If it reads almost zero, the two coils are opposing each other. Turn the third pair around and repeat.
~1/10 – 1/12??VYWZVV ≠ 0
Correct: with Y joined to W, the two coils in series add their voltages.
~1/10 – 1/12??VYZWVV ≈ 0
Reversed: with Y joined to Z, the two voltages cancel out.

Step 3: second test

  1. Now feed V-Y, one of the two coils you have already marked.
  2. Take W-Z as good: it was not fed in the first test and already has its letters. Join Z to one lead of the remaining pair and put the voltmeter between W and the other lead.
  3. If there is a clear voltage, the lead joined to Z is U and the free one is X. If it reads almost zero, turn that pair around and repeat.
~1/10 – 1/12VYWZUXVV ≠ 0
With W-Z as the reference, the coil fed in the first test is determined.

Check

Feed W-Z, join X to V and measure between U and Y. You should see a clear voltage, similar to the earlier ones. The three joints used are now end to start: X-V, Y-W and Z-U.

First run

Connect the motor in star or delta according to its nameplate, start it with no load and measure the three currents: they should be similar to one another. If they differ widely or the motor growls, stop and repeat the tests.

The letters are relative: the three coils are consistent with one another, but the one you called U need not be the factory U. If it turns the wrong way, swap two supply phases; do not change the letters.

Why it works

The three phases are spaced 120 electrical degrees apart. The flux of the fed coil links the other two equally and induces the same voltage in both, with the same sign when each is followed from start to end. So, in series end to start, they add. With one reversed, they subtract and almost cancel:

\[ E_V + E_W \neq 0 \qquad\qquad E_V - E_W \approx 0 \]

We checked the sign using the real slot distribution of five windings, from 12 to 48 slots. Do not expect to read the applied voltage: the rotor cage opposes the flux and the reading usually stays below it. What matters is the difference between a clear reading and a near-zero one, and a small remainder in the near-zero reading is normal.

If something does not add up

What you seeWhat to check
Low reading in both positionsThe source is collapsing or too small. Measure the voltage on the fed coil.
Clear reading in both positionsThe pairs are wrongly separated, or there is contact between coils or to the frame.
The check gives no voltageA letter is wrong. Start again from the first test.
Very different readings between testsPossibly a faulty coil, such as shorted turns or a different number of turns. The winding should be inspected.

Related guides